Gaffaweb >
Love & Anger >
1992-28 >
[ Date Index |
Thread Index ]
[Date Prev] [Date Next] [Thread Prev] [Thread Next]
From: v129j6ed@ubvmsb.cc.buffalo.edu (This is F.U.N.)
Date: Tue, 29 Sep 1992 15:21:00 GMT
Subject: Re: Post Results from the All-Out Kate Bush Poll
News-Software: VAX/VMS VNEWS 1.41
Newsgroups: rec.music.gaffa
Organization: University at Buffalo
References: <9209251223.AA05602@relay1.UU.NET>
Sender: nntp@acsu.buffalo.edu
In article <9209251223.AA05602@relay1.UU.NET>, DKASTENS@dosuni1.rz.Uni-Osnabrueck.DE (Dirk Kastens) writes... >Greetings, >Thanks very much for the poll. >But I think, the result needs some interpretation. >A great problem in statistical analysis are the missing values. If you >make an interview and some people don't respond to a question, how can >you rate the result? >In the Kate poll you didn't care about the missing values, you just >calculated the mean (sum of points / number of values). Is this >correct? >Let's examine one extreme case: >song A = 1 x 10 points, 35 x missing value -> mean: 10/1 = 10 >song B = 35 x 10 points, 1 x 9 points -> mean: 359/36 = 9.9722 >Which one is better? >Regarding the albums we find Vol.1 before TSW. But only half of the >voters gave out points for Vol.1, whereas all of them valued TSW. The same >with the new number one song Under the Ivy. >I think that it's more fair to order the songs by the sum of points. >What do you think? I always found that knowledge is power. The fact is that if someone does not know a song, why should that harm it's average. The truth is we don't know how someone will vote for it. The assumption I went under was that total points mean nothing. Let's use another extreme example: song A = 300 points, 60 votes song B = 298 points 30 votes This could be disadvantageous to a song voted 1 more time than another. "Suspended in Gaffa" and the "Jig of Life" would have tied for the top. The whole poll, from my point of view, was what people thought about each song. That came out pretty well since all the songs were voted for over 60% of the time. Using statistical method would have been better, but since I am a History major, I am not sure how to set this up. ******************************************************************************* "As your Mr.Spock said, The needs of the many outweighed the needs of the few." "A truism, like most of Spock's Epigrams. But on occasion, I'm afraid, cold comfort." Jillian and James T. Kirk - "Star Trek - Debt of Honour" ******************************************************************************* Jef Kolodziej The KaTeFan(tm) v129j6ed@ubvms.cc.buffalo.edu State University of New York at Buffalo History Major 51 East NorthRup (716) 837-9610 Buffalo, NY 14214